Floating-point numbers are stored as string types before calculation (I don't know whether they are true or false).

When floating-point data is very large or small (six bits can be moved back and forth), floating-point data will be represented by e notation, and the maximum storage capacity is 17 bits. In addition, the isFinite () method can judge whether a numerical value is finite, and the isNaN () method can judge whether a data is a non-numerical type. Other types can be converted to numeric types by parseInt () and parseFloat (), and NaN (not a number) is returned if the conversion is unsuccessful.